Continues Target Tracking Based On Ncc And Kalman Algorithm

When we use Kalman tracking algorithm to track objects, the tracker will fail if the object tends to stopping. This paper proposed a template matching method based on NCC and Kalman tracking algorithm that solved this problem. We save the template and center point of the last frame, when the object stops or tends to be static then we can use the saved template and center point to search the object in the nearby area, then we can get the actual position of the object we are tracking. The experimental results showed that: the proposed method can keep tracking the target by switching to the improved NCC template matching algorithm when the target tends to stop, also we attain the higher tracking accuracy and smaller center of errors than Meanshift by getting the most matching region.